Version 1.0, August 31, 2001, Copyright, Hugh Jack 1993-2001
An interesting approach to path planning was suggested by B.Espiau and R.Boulic [1986]. In the attempt to find a path planning method for a 10 degree of freedom manipulator, they came up with an unusual approach to optimization. With proximity or contact sensors mounted on every link of the manipulator, the path was navigated. To do this the sensor data would be used alter the penalty functions to represent detected obstacles encountered by the manipulator. The cost function was based on velocity and dynamic functions. This gave a dynamic approach to finding paths in which the path was found by the local optimization of trajectories. For this particular method a large number of sensors are required, and unfortunately the authors did not provide statistics about the performance of the method.